Testing times for Derek Lowe

Drug discovery chemists live by assay data; we depend on these numbers to tell us if we’re heading in the right direction with our molecules. Ideally, every major factor that goes into selecting a clinical candidate will have an assay associated with it. We expect the assays at the front end of a project to cut some wide paths through our compounds - most things don’t work, after all. No one gets too worked up about the data from the primary screen, as long as the numbers make some sort of sense. But there are other assays, further along the process, that no one enjoys facing.

There are several reasons for an assay to end up on the ’most feared’ list. Some are there for their consequences - they rarely return anything bad, but when they do, your compound may well be dead. The Ames test is one of those. Using a panel of mutated bacteria that have trouble repairing damage to their DNA, the Ames test looks to see if your compound is causing any. A positive Ames isn’t all that common, but it will cast an immediate shadow over a drug candidate in most organisations. Why take the chance of developing an Ames-positive compound, when you could back up and try to find something better? ’It might not cause mutations’ is not much of a marketing slogan.

Other tests are worth worrying about because it’s hard to know what to do about their results. Induction or inhibition of cytochrome enzymes are in this category. You don’t want your drug to throw off the normal levels of liver enzymes - unless you’re sure, for some reason, that your patient population will never be taking anything else while they’re taking your drug. (And who’s ever sure about that?) Compounding the problem is that, as with the Ames test, the reasons for a worrisome result may not be very clear. All sorts of compounds interact with the metabolising enzymes - their whole purpose is to recognise a huge variety of foreign substances. So trying to outwit them, while possible, is never easy. Your best hope is to run as many structurally related compounds through the same assay, hope for a trend to emerge, and hope that it doesn’t take you away from the activity you wanted in the first place.


Those assays share another difficulty with others on the list: it can be hard to know where to draw the line with them. A huge effect on liver enzymes is clearly a killing blow to a drug candidate. But what if it’s a more moderate one, perhaps just in one enzyme subtype? How much should you worry then? Another of the fearsome assays is notorious for this problem: the hERG test. That, regrettably, is an acronym for the ’human ether-a-go-go related gene’, whose bizarre name can be traced (like so many others) to the rather loosely-wound fruit-fly geneticists. It’s an ion channel in the heart, and ligands for it can lead to dangerous arrhythmias. Several drugs have already been pulled from development (or even from pharmacy shelves) due to hERG problems, so drug developers take it very seriously indeed. 

But the problem is that no one completely trusts the in vitro assays for it. There are several different ways to run them, which can give you different results, and all of them are notorious for producing ’sort-of kind-of’ data. It’s rare to see a compound strongly light up these assays, but when one does, it’s immediately dropped. It’s distressingly common to see lead compounds show a bit of activity here and there, and that’s enough to make everything that much more difficult and expensive. Extra cardiovascular tests get ordered in the rodents, then in the dogs, and finally in the human volunteers if the compound survives into the clinic.

And for people who work in preclinical drug development, there’s one assay that can never be taken lightly: two-week rat toxicity testing. It comes near the end of the project and it’s often the first real, in-depth look at the side effects of a compound beyond what can be noted in a single-dose experiment. By this time you’ve narrowed down to a very short list of candidates, with a lot of time and effort invested in each of them. Any project leader who is unconcerned about this stage of testing has never been through it before.

Wait for the histopathologists 

That’s because our knowledge of real whole-animal toxicology makes the other assays I’ve discussed look wide-open and friendly. Two continuous weeks of a good strong dose of your drug candidate can expose all sorts of unusual things, the reasons for which may never be satisfactorily understood. Even if the rats make it through the test with no noticeable problems, experienced team members know to wait for the histopathologists to weigh in before starting to breath normally again. If a compound can make it past a full microscopic examination of a dozen different tissues, then it’s definitely something to take seriously - and congratulations are in order. 

Derek Lowe is a medicinal chemist working on preclinical drug discovery in the US.